"""Wrapper around Together AI's Completion API.""" import logging from typing import Any, Dict, List, Optional from aiohttp import ClientSession from langchain_core._api.deprecation import deprecated from langchain_core.callbacks import ( AsyncCallbackManagerForLLMRun, CallbackManagerForLLMRun, ) from langchain_core.language_models.llms import LLM from langchain_core.pydantic_v1 import Extra, SecretStr, root_validator from langchain_core.utils import convert_to_secret_str, get_from_dict_or_env from langchain_community.utilities.requests import Requests logger = logging.getLogger(__name__) @deprecated( since="0.0.12", removal="0.3", alternative_import="langchain_together.Together" ) class Together(LLM): """LLM models from `Together`. To use, you'll need an API key which you can find here: https://api.together.xyz/settings/api-keys. This can be passed in as init param ``together_api_key`` or set as environment variable ``TOGETHER_API_KEY``. Together AI API reference: https://docs.together.ai/reference/inference """ base_url: str = "https://api.together.xyz/inference" """Base inference API URL.""" together_api_key: SecretStr """Together AI API key. Get it here: https://api.together.xyz/settings/api-keys""" model: str """Model name. Available models listed here: https://docs.together.ai/docs/inference-models """ temperature: Optional[float] = None """Model temperature.""" top_p: Optional[float] = None """Used to dynamically adjust the number of choices for each predicted token based on the cumulative probabilities. A value of 1 will always yield the same output. A temperature less than 1 favors more correctness and is appropriate for question answering or summarization. A value greater than 1 introduces more randomness in the output. """ top_k: Optional[int] = None """Used to limit the number of choices for the next predicted word or token. It specifies the maximum number of tokens to consider at each step, based on their probability of occurrence. This technique helps to speed up the generation process and can improve the quality of the generated text by focusing on the most likely options. """ max_tokens: Optional[int] = None """The maximum number of tokens to generate.""" repetition_penalty: Optional[float] = None """A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition. """ logprobs: Optional[int] = None """An integer that specifies how many top token log probabilities are included in the response for each token generation step. """ class Config: """Configuration for this pydantic object.""" extra = Extra.forbid @root_validator(pre=True) def validate_environment(cls, values: Dict) -> Dict: """Validate that api key exists in environment.""" values["together_api_key"] = convert_to_secret_str( get_from_dict_or_env(values, "together_api_key", "TOGETHER_API_KEY") ) return values @property def _llm_type(self) -> str: """Return type of model.""" return "together" def _format_output(self, output: dict) -> str: return output["output"]["choices"][0]["text"] @staticmethod def get_user_agent() -> str: from langchain_community import __version__ return f"langchain/{__version__}" @property def default_params(self) -> Dict[str, Any]: return { "model": self.model, "temperature": self.temperature, "top_p": self.top_p, "top_k": self.top_k, "max_tokens": self.max_tokens, "repetition_penalty": self.repetition_penalty, } def _call( self, prompt: str, stop: Optional[List[str]] = None, run_manager: Optional[CallbackManagerForLLMRun] = None, **kwargs: Any, ) -> str: """Call out to Together's text generation endpoint. Args: prompt: The prompt to pass into the model. Returns: The string generated by the model.. """ headers = { "Authorization": f"Bearer {self.together_api_key.get_secret_value()}", "Content-Type": "application/json", } stop_to_use = stop[0] if stop and len(stop) == 1 else stop payload: Dict[str, Any] = { **self.default_params, "prompt": prompt, "stop": stop_to_use, **kwargs, } # filter None values to not pass them to the http payload payload = {k: v for k, v in payload.items() if v is not None} request = Requests(headers=headers) response = request.post(url=self.base_url, data=payload) if response.status_code >= 500: raise Exception(f"Together Server: Error {response.status_code}") elif response.status_code >= 400: raise ValueError(f"Together received an invalid payload: {response.text}") elif response.status_code != 200: raise Exception( f"Together returned an unexpected response with status " f"{response.status_code}: {response.text}" ) data = response.json() if data.get("status") != "finished": err_msg = data.get("error", "Undefined Error") raise Exception(err_msg) output = self._format_output(data) return output async def _acall( self, prompt: str, stop: Optional[List[str]] = None, run_manager: Optional[AsyncCallbackManagerForLLMRun] = None, **kwargs: Any, ) -> str: """Call Together model to get predictions based on the prompt. Args: prompt: The prompt to pass into the model. Returns: The string generated by the model. """ headers = { "Authorization": f"Bearer {self.together_api_key.get_secret_value()}", "Content-Type": "application/json", } stop_to_use = stop[0] if stop and len(stop) == 1 else stop payload: Dict[str, Any] = { **self.default_params, "prompt": prompt, "stop": stop_to_use, **kwargs, } # filter None values to not pass them to the http payload payload = {k: v for k, v in payload.items() if v is not None} async with ClientSession() as session: async with session.post( self.base_url, json=payload, headers=headers ) as response: if response.status >= 500: raise Exception(f"Together Server: Error {response.status}") elif response.status >= 400: raise ValueError( f"Together received an invalid payload: {response.text}" ) elif response.status != 200: raise Exception( f"Together returned an unexpected response with status " f"{response.status}: {response.text}" ) response_json = await response.json() if response_json.get("status") != "finished": err_msg = response_json.get("error", "Undefined Error") raise Exception(err_msg) output = self._format_output(response_json) return output